The 2nd Perinatal HIV Prevention Trial (PHPT-2) in Thailand showed that a single dose of nevirapine (SD- NVP) to women at onset of labor and to infants at 48-72 hours in addition to zidovudine (ZDV) starting at 28 weeks'gestation, could reduce HIV perinatal transmission to about 2%. However, post-exposure NVP resistance mutations detected in mothers have raised concerns regarding future efficacy of Non Nucleoside Reverse Transcriptase Inhibitor based treatments for their own health. Results from a study in Botswana suggested that maternal SD-NVP may not be needed if the infant SD-NVP is administered immediately after birth. The study primary objective is to test whether maternal SD-NVP can be omitted without loss of efficacy in preventing perinatal HIV transmission. A regimen with no maternal NVP and two infant NVP doses, one immediately after birth, the other after 48-72 hours, will be compared for efficacy with a regimen including a maternal NVP dose during labor and identical infant NVP dosing. ZDV prophylaxis will be provided from 28 weeks gestation or as soon as possible thereafter, loading dose during labor and 1 week in the infants. All infants will be formula fed. To diminish the risk of selecting NVP resistance mutations, mothers exposed to NVP will receive a one week course of ZDV plus lamivudine, following WHO and Thai guidelines. The secondary objectives are to study HIV resistance mutations to NVP in HIV infected infants, to compare the safety of both strategies in mothers and infants and to study the pharmacokinetics of NVP in infants. This multicenter, phase III, randomized, double-blind, controlled trial will enroll 1398 consenting HIV-1 infected pregnant women, with CD4 count >200/mm3, participating in the Thai Ministry of Public Health (MoPH) prevention program, assigned to one of the 2 study arms (699 per arm including 5% non evaluable;power: 80%;delta threshold for non-inferiority testing between groups: 1.5%). The primary endpoint is infant HIV status based on confirmed DNA-PCR results on specimens drawn at birth, 1, 2, 4 and 6 months. The study will build upon PHPT-1 and PHPT-2, a collaborative effort between Harvard University, IRD- France, the Thai MoPH, and Mahidol and Chiang Mai Universities, through a network of 39 public hospitals. If maternal NVP proves unnecessary, a simple regimen will be available for the prevention of perinatal HIV transmission, resistance mutations avoided, and maternal future drug options preserved. This study will have important public health implications in settings where NVP, with or without ZDV background, is a central component of the prevention of intrapartum transmission, It will also provide key scientific knowledge on the role of post-exposure prophylaxis in the prevention of intrapartum HIV transmission.